Dear DAVID Community,

As you are probably aware, the entire DAVID company, team and product range has been integrated into HP Inc.
This forum is now read-only. Relevant content has been migrated to HP and merged into HP's Support Forums on November 1, 2016.

To start new discussions for 3D Scanning, please register and post your new topic at the HP Support Forums

Scanning of calibration panel

The place for questions, problems, comments and tips regarding the camera calibration.

Re: Scanning of calibration panel

Postby MagWeb » Wed Dec 11, 2013 2:12 am

:? I 'd say: You did a good job! And: Thanks!

The lower the frequency the better you can get cavities - this fits to my experience.
But your results made me wonder too:
Your frequency5 result @ 800*600 shows 2 directions of waves. What the hell causes this second direction at a lower frequency (I saw this before here too, but didn`t care about)? I'm not through yet...seems to be another problem of balance...
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Scanning of calibration panel

Postby Sven » Wed Dec 11, 2013 11:30 am

Thank you, argo and Gunter, for your constructive dialog!

Yes, it can always be difficult to use "other" hardware that has not been tested/recommended. The camera resolution is often important. You never know what they do inside the camera...
In earlier days, for laserscanning, we often found that the best resolutions are the native chip resolution or exactly half. Seems to be the same here (1280*720 is native, 960*720 probably just cuts left+right off).

The other important topic is frame rate, but it seems you two have solved that quickly. :D

Sven
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: Scanning of calibration panel

Postby argo » Wed Dec 11, 2013 12:47 pm

MagWeb wrote:The lower the frequency the better you can get cavities - this fits to my experience.

Agree. But sometimes also there will be some unwanted (Unreal Shapes) using the lower frequencies. but in total i would say the lower is better (not too much low)
The question is why the frequency value, backs to default after each scan? it is so boring to set it for each and every scan!!

seems to be another problem of balance...

I did not get what you mean!. is it related to my side or David's side?

Sven wrote:The other important topic is frame rate, but it seems you two have solved that quickly

Yup and thanks to Gunter for that :)

There is a question and maybe not related to this topic but still calibration matter:
Assuming i have a 240mm Scaled panel and my object is 200mm tall.
What would be the Calibr.Scale value in David? 240mm? 200mm?
Or i have to print a separate calibration for this object?

Regards,
User avatar
argo
 
Posts: 141
Joined: Sun Jul 14, 2013 8:48 am

Re: Scanning of calibration panel

Postby Sven » Wed Dec 11, 2013 1:06 pm

Frequencies: Please disable "AutoFrequencies" directly above "Frequencies".
Scale: Of course that's the scale of the calibration panel. See manual. http://wiki.david-3d.com/david3_user_manual/overview
Image
For a 200 mm object, the 240mm calibration panel should be fine.
Sven
User avatar
Sven
DAVID Developer
 
Posts: 1541
Joined: Mon Jan 22, 2007 8:59 am
Location: Braunschweig, Germany

Re: Scanning of calibration panel

Postby argo » Wed Dec 11, 2013 2:42 pm

Sven wrote:Frequencies: Please disable "AutoFrequencies" directly above "Frequencies".

Thank you Sven.. how i missed that!?
Sven wrote:For a 200 mm object, the 240mm calibration panel should be fine.

The Result is about 170 mm. that is why i asked to be sure about the scale.
I have to look and check every possible things that caused this mismatched size!!
But what are them?!
For small objects i have three different calibration panels: (scales of: 30 mm, 60mm, 90 mm) and the results are exact size of real one(sometimes maybe some millimeters mismatches)
But in case of using 240 mm scaled panel for object's tall between 200 mm and 500 mm, i could not see any exact sized scans (except for object's tall same or close to 240 mm).

And it is fine with me if we need to print different calibration scale for each object to have a perfect scan size.
User avatar
argo
 
Posts: 141
Joined: Sun Jul 14, 2013 8:48 am

Re: Scanning of calibration panel

Postby MagWeb » Wed Dec 11, 2013 3:42 pm

argo wrote:But in case of using 240 mm scaled panel for object's tall between 200 mm and 500 mm, i could not see any exact sized scans (except for object's tall same or close to 240 mm).

Strange thing:
You are using a "negative" corner, right?
Maybe the pattern offset value (advanced settings) is the bad guy? Measure the distance between the markers next to the corners edge (ideal: it should be = Scale/8). Offset = measured distance - Scale/8
Just a guess

In addition to Sven about Scale in general:
The Scale value simply tells DAVID how big the calibration target is in our real world.
You could also put in a value in inches as long as your ruler was in inches, and your 3D object's size should be 1:1 in inches.
You can also use the Scale to influence the output size. Many 3D apps do not allow very big dimensions at their default settings. So if you plan to use your scan in such a app (let's say its default scene is limited to a box of 100x100x100mm but your object is actually 1000mm) you may put in a Scale value of MeasuredValue/10. Your final 3D object would be scaled down now to 1/10 and fits to the app's scene.
(note: a few parameters in DAVID, as the BackgroundFilterFactor or the result filtering clipping planes depend on this Scale value and need to be in the same ratio if used)

argo wrote: "seems to be another problem of balance..."
I did not get what you mean!. is it related to my side or David's side?

Most likely this is a problem on the side of the user's settings. Setting up the different compenents the right, balanced way is mostly based on the result-experiences of the user. Only in rare cases he actualy knows why his result got better or worse. I just try to find the different problems. The second wave direction in your frequ5 result seems to be caused by something different than we talked about above...
You may agree that's a hard way to get an individual system to optimal results. We could ease this way knowing about its different glitches and develop a concrete step by step workflow to avoid them.
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Scanning of calibration panel

Postby argo » Wed Dec 11, 2013 7:21 pm

MagWeb wrote:You are using a "negative" corner, right?

Yes. you are right.

MagWeb wrote:Maybe the pattern offset value (advanced settings) is the bad guy? Measure the distance between the markers next to the corners edge (ideal: it should be = Scale/8). Offset = measured distance - Scale/8

The result is zero and true. (the distance between center of markers are exactly 60 mm which for 240 mm is correct!?)
I am so sure about everything related to my calibration panel.

MagWeb wrote:The second wave direction in your frequ5 result seems to be caused by something different than we talked about above...

Agree. but did not care about it. what can i do to find out the reason?

MagWeb wrote:You may agree that's a hard way to get an individual system to optimal results. We could ease this way knowing about its different glitches and develop a concrete step by step workflow to avoid them.

So true. if there is anything i can do, just let me know.

Regards,
User avatar
argo
 
Posts: 141
Joined: Sun Jul 14, 2013 8:48 am

Re: Scanning of calibration panel

Postby MagWeb » Thu Dec 12, 2013 3:47 am

What's the current parameter at AdvanceSettings/Calibration/ExpectedPatternsOffset (or named similar... I'm of my PC atm...default is about 15, if I remember right)? If your pattern's offset is equal zero you should set zero instead.
As said, just a guess.

Other things tomorrow...
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Previous

Return to Camera Calibration

Who is online

Users browsing this forum: No registered users and 2 guests

cron